The overall goal of this Mentored Patient-Oriented Research Career Development (K23) Award is to 1. Provide the applicant with the skills to perform independent functional magnetic resonance brain imaging (fMRI) research in Eating Disorders, specifically Restricting Type Anorexia Nervosa (RAN), 2. Train the applicant in novel Neurocomputational Modeling approaches that link psychological theory with underlying neural substrates, and 3. Extend our knowledge on brain activity in RAN in relation to food and non-food (monetary) reward processing. Aim 1 .a. is to investigate the functionality of anteroventral striatum reward pathways using fMRI and a paradigm of expected and unexpected delivery of sweet taste stimuli (O'Doherty et al 2003). We hypothesize that women Recovered from and III with RAN will have reduced anteroventral striatum activation compared to control women. In Aim 1.b., using computational models for reward conditioning on the fMRI data, we expect reduced reward learning signals in III and Recovered RAN in the anteroventral striatum compared to controls. Aim 2.a. is to investigate if reduced orbitofrontal cortex and striatum activation compared to CW exists in III and Recovered RAN in response to monetary immediate reward (Tanaka et al 2004). In contrast we expect III and recovered RAN to have greater activation than controls in the dorsolateral prefrontal and parietal cortex during delayed monetary reward (Tanaka et al 2004). In Aim 2.b., using computational models on the acquired brain data, we expect in III and Recovered RAN reduced anteroventral striatum immediate, but increased dorsal striatum delayed reward processing compared to controls. This study uses novel technologies and paradigms. It will both advance the field of Eating Disorders research greatly as well as will be an excellent opportunity to provide the applicant with training in fMRI and Neurocomputational Modeling. The applicant will be mentored by renowned experts in the field of Eating Disorders, fMRI Brain Imaging and Neurocomputational Modeling, and he will attend courses in the Cognitive Science and Psychology Program at the University of Colorado. Anorexia Nervosa has the highest mortality among all psychiatric disorders. This application will help identify pathways of disturbance of the reward system in Restricting type AN, will provide new insight into the disease mechanism, and will provide us with new information for the development of advanced treatments.